net/sfc: support event queue

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc: interrupts support sufficient for event queue init

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc: estimate available resources

Resources required in accordance with configuration are
allocated only.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc: add device start and stop operations

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc: add device configuration checks

Manual link speed/duplex configuration is not supported yet.
Loopback is not supported yet.
Flow Director is not supported.
Link status change notification using interrupt is not supported yet.
Receive data notification using interrupts is not supported yet.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc: add configure and close stubs

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc: add init on attach

The setup and configuration of the PMD is not performance sensitive,
but is not thread safe either. It is possible that the multiple
read/writes during PMD setup and configuration could be corrupted
in a multi-thread environment.  Since this is not performance
sensitive, the developer can choose to add their own layer to provide
thread-safe setup and configuration. It is expected that, in most
applications, the initial configuration of the network ports would be
done by a single thread at startup.

In the case of exception on the event queue, the event queue and
corresponding Rx/Tx queue should be restarted in the Rx/Tx queue
polling context. These operations require access to the device
control which should be serialized. The device level lock will do
the job.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc: build libefx

Implement efsys.h for the PMD.

Signed-off-by: Artem Andreev <artem.andreev@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import libefx licensing

Provide API to deal with licenses on SFN7xxx and SFN8xxx
family adapters.

EFSYS_OPT_LICENSING should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import bootrom configuration

Provide API to read/write bootrom configuration from/to NVRAM.

EFSYS_OPT_BOOTROM should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import VPD support

Provide API to read/write PCI Vital Product Data.

EFSYS_OPT_VPD should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import NVRAM support

Provide API to work with NIC non-volatile memory. It is used
to update firmware, configure NIC including bootrom parameters,
manage licenses, store PCI Vital Product Data etc.

EFSYS_OPT_NVRAM should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import Rx packed stream mode

In packed stream mode, large buffers are provided to the NIC
into which many packets can be delivered. This reduces the
number of queue refills needed compared to delivering every
packet into a separate buffer.

EFSYS_OPT_RX_PACKED_STREAM should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import monitors access via MCDI

EFSYS_OPT_MON_MCDI should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import monitors statistics

EFSYS_OPT_MON_STATS should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import loopback control

EFSYS_OPT_LOOPBACK should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import RSS support

EFSYS_OPT_RX_SCALE should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import Rx scatter support

EFSYS_OPT_RX_SCATTER should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import event prefetch

EFSYS_OPT_EV_PREFECT allows to enable event prefetching
when event queue is polled.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import MAC statistics

MAC statistics are either periodically (if supported/requested)
or on-demand written to provided DMA-mapped memory.
If periodic update is not supported (e.g. for EF10 virtual
functions), it is the driver responsibility to handle it.

EFSYS_OPT_MAC_STATS should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import PHY LEDs control

EFSYS_OPT_PHY_LED_CONTROL should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import PHY statistics

EFSYS_OPT_PHY_STATS should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import PHY flags control

EFSYS_OPT_PHY_FLAGS should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import software per-queue statistics

EFSYS_OPT_QSTATS should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import built-in selftest

EFSYS_OPT_BIST should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import diagnostics support

EFSYS_OPT_DIAG should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import SFN8xxx family support

SFN8xxx is the second family based on EF10 architecture.

It has few differences from SFN7xxx adapters family.

EFSYS_OPT_MEDFORD should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import SFN7xxx family support

SFN7xxx is the first family based on EF10 architecture.

EFSYS_OPT_HUNTINGTON should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import 5xxx/6xxx family support

EFSYS_OPT_SIENA should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import MCDI proxy authorization

MCDI proxy authorization may be used if privileged PCI
function (physical function) would like to intercept and
authorize MCDI requests done by unprivileged (e.g. virtual)
PCI function. It may be used to control unprivileged
function Rx mode (e.g. promiscuous, all-multicast), MTU
and default MAC address change requests etc.

Current libefx support is limited to client-side which
is required to work when function requests need to be
authorized.

Server side support required to request and do the
authorization is not implemented yet.

EFSYS_OPT_MCDI_PROXY_AUTH should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import MCDI logging

Driver can provide a function to be called to log MCDI
requests and responses to help with debugging.

Solarflare netlogdecode cross-platform tool may be used
to decode these logs.

EFSYS_OPT_MCDI_LOGGING should be enabled to use it.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import MCDI implementation

Implement interface to talk to NIC management CPU. Provide
helpers to fill in MCDI requests, execute it and process
received response.

MCDI request is prepared in either PCI BAR mapped memory
(SFN5xxx/SFN6xxx) or DMA-mapped memory (SFN7xxx/SFN8xxx) and,
doorbell is pressed (memory-mapped register) to execute it.

Events about MCDI completion are delivered to house-keeping
event queue, but usage of these events is optional and MCDI
buffer may be simply polled waiting for completion
indication set.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import MCDI definition

The header defines data interface between host CPU and NIC
management CPU.

The header is automatically generated from firmware sources.

MCDI is used on NIC control path (configuration,
event/transmit/receive queues setup and teardown etc), but
not used on data path.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import filters support

Filtering capabilities depend on NIC family and used firmware
variant. Provided API allows to get supported filter types
(in a priority order), add/delete individual filters and
restore entire filter table after, for example, NIC management
CPU reboot.

Rx filters allow to redirect matching flow to specified Rx queue.

Tx filters allow to control generated traffic (e.g. to implement
virtual function anti-spoofing control).

EFSYS_OPT_FILTER should be enabled to use it. It is required
for SFN7xxx and SFN8xxx adapter families support.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import register definitions

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc/base: import libefx base

libefx is a platform-independent library to implement drivers
for Solarflare network adapters. It provides unified adapter
family independent interface (if possible).

Driver must provide efsys.h header which defines options
(EFSYS_OPT_*) to be used and macros/functions to allocate
memory, read/write DMA-mapped memory, read/write PCI BAR
space, locks, barriers etc.

efx.h and efx_types.h provide external interfaces intended
to be used by drivers. Other header files are internal.

From Solarflare Communications Inc.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/sfc: libefx-based driver stub

Enable the PMD by default on supported configurations.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/i40evf: fix casting between structs

Casting from structs which lay out data in typed members
to structs which have flat memory buffers, will cause
problems if the alignment of the former isn't as expected.
This patch removes the casting between structs.

Fixes: ae19955e7c86 ("i40evf: support reporting PF reset")

Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>

net/e1000/base: announce supported devices

Document all supported NICs.
Add Intel I219 NICs support in release note.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/i40e: enable auto link update for 25G

For 25G devices auto link update was disabled because it was causing
link issues when enabled.

The problem found because of interface changes in admin queue command
"set_phy_config" and "get_phy_capabilities" for 25G.

This patch fixes the issue and enables auto link update for 25G devices.

Fixes: 75d133dd3296 ("net/i40e: enable 25G device")

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>

net/i40e: fix logging for Tx free threshold check

Fixes: 4861cde46116 ("i40e: new poll mode driver")

Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>

net/e1000: enable new I219 devices

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: update shared code version

Updated to 2016.11.22

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: support more I219 devices

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: disable force K1-off feature

MAC-PHY desync may occur causing misdetection of link up event.
Disabling K1-off feature can work around the problem.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: add workaround for possible stalled packet

This works around a possible stalled packet issue, which may occur due to
clock recovery from the PCH being too slow, when the LAN is transitioning
from K1 at 1G link speed.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: enable new I219 devices

Enable the support of new I219 devices.
Also define some registers for future usage.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: add workaround for ULP entry flow

For I217 revision 6, when entering Ultra Low Power (ULP) we need to enable
Low Power Link Up (LPLU) and disable Gig speed to make it work.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: increase LANPHYPC low duration

LANPHYPC low duration of 10 usec was too low for some corner cases
causing interface mismatches during Ultra Low Power (ULP) exit.

This patch increases the LANPHYPC low duration to 1 msec.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: clear ULP configuration register on ULP exit

There are some client PHY Ultra Low Power (ULP) register bits that are
configured by the Manageability Engine (ME) FW.

The driver must ensure that these bits are cleared on exit from ULP.
Ordinarily the ME FW would do that, but there are cases in which the
FW is not present, and the driver must handle that.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: restore link speed after ULP exit

When Ultra Low Power (ULP) enabled, the client PHY needs to be set up
for link configuration after cable reconnected.

Previously link configuration was only done in auto-negotiate mode.

Do link configuration also in autoneg disabled mode.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: define max Rx jumbo frame size

Add definition MAX_RX_JUMBO_FRAME_SIZE for igb.

All igb parts (82575 and newer) have 9.5K max jumbo frame size.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: expose I350 internal function

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: get FW version for I354

I354 support was missing in the e1000_get_fw_version() which resulted in
the FW version not being reported. Support added.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: retry to get HW mailbox lock

The driver shouldn't give up if it fails to get the hardware mailbox lock.

This can happen in a situation where the PF-VF communication channel is
heavily loaded and causes complete communications failure between the PF
and VF drivers.

Add a counter and a delay. The driver will now retry ten times,
waiting one millisecond between retries.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: avoid packet loss for non-1G

To avoid packet loss, Phase Lock Loop (PLL) clock gate time needs to be
increased for non 1G speeds.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/e1000/base: increase ULP timer

With new hardware (I219), Ultra Low Power (ULP) exit takes significantly
longer time. Therefore, driver must wait longer.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net: align ethdev and eal driver names

Some virtual pmds report a different name than the vdev driver name
registered in eal.
While it does not hurt, let's try to be consistent.

Signed-off-by: David Marchand <david.marchand@6wind.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net: remove dead driver names

Since commit b1fb53a39d88 ("ethdev: remove some PCI specific handling"),
rte_eth_dev_info_get() relies on dev->data->drv_name to report the driver
name to caller.

Having the pmds set driver_info->driver_name in the pmds is useless,
since ethdev overwrites it right after.
The only thing the pmd must do is:
- for pci drivers, call rte_eth_copy_pci_info() which then sets
  data->drv_name
- for vdev drivers, manually set data->drv_name

At this stage, virtio-user does not properly report a driver name (fixed in
next commit).

Signed-off-by: David Marchand <david.marchand@6wind.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Reviewed-by: Jan Blunck <jblunck@infradead.org>

net/mlx5: do not invalidate title CQE

We can leave the title completion queue entry untouched since its contents
are not modified.

Reported-by: Liming Sun <lsun@mellanox.com>
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

net/mlx5: fix endianness in Tx completion queue

Completion queue entry data uses network endian, to access them we should
use ntoh*().

Fixes: c305090bbaf8 ("net/mlx5: replace countdown with threshold for Tx completions")

Reported-by: Liming Sun <lsun@mellanox.com>
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

net/mlx5: fix leak when starvation occurs

The list of segments to free was wrongly manipulated ending by only freeing
the first segment instead of freeing all of them.  The last one still
belongs to the NIC and thus should not be freed.

Fixes: a1bdb71a32da ("net/mlx5: fix crash in Rx")

Reported-by: Liming Sun <lsun@mellanox.com>
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

net/i40e: fix spelling

Fixes: da61cd084976 ("i40evf: add extended stats")
Fixes: 0eedec25ea36 ("i40e: clean log messages")

Signed-off-by: Remy Horton <remy.horton@intel.com>
Acked-by: Kevin Traynor <ktraynor@redhat.com>

net/i40e: fix xstats value mapping

The offsets used in rte_i40evf_stats_strings for transmission
statistics were wrong, returning the total byte count rather than
the respective (unicast, multicast, broadcast, drop, & error)
packet counts.

Fixes: da61cd084976 ("i40evf: add extended stats")

Signed-off-by: Remy Horton <remy.horton@intel.com>
Acked-by: Kevin Traynor <ktraynor@redhat.com>

net/virtio: fix build without virtio-user

When CONFIG_RTE_VIRTIO_USER is disabled (default on FreeBSD),
the virtio driver cannot be compiled:

librte_pmd_virtio.a(virtio_ethdev.o): In function `eth_virtio_dev_init':
(.text+0x1eba): undefined reference to `virtio_user_ops'

Reported-by: Andrew Rybchenko <arybchenko@solarflare.com>
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>

examples/ethtool: display firmware version

This patch enhances the ethtool example to support to show
firmware version, in the same way that the Linux kernel
ethtool does.

Signed-off-by: Qiming Yang <qiming.yang@intel.com>
Acked-by: Remy Horton <remy.horton@intel.com>

net/i40e: add firmware version get

This patch add a new function i40e_fw_version_get.

Signed-off-by: Qiming Yang <qiming.yang@intel.com>
Acked-by: Remy Horton <remy.horton@intel.com>

net/ixgbe: add firmware version get

This patch adds a new function ixgbe_fw_version_get.

Signed-off-by: Qiming Yang <qiming.yang@intel.com>
Acked-by: Remy Horton <remy.horton@intel.com>

net/e1000: add firmware version get

This patch adds a new function eth_igb_fw_version_get.

Signed-off-by: Qiming Yang <qiming.yang@intel.com>
Acked-by: Remy Horton <remy.horton@intel.com>

ethdev: add firmware version get

This patch adds a new API 'rte_eth_dev_fw_version_get' for
fetching firmware version by a given device.

Signed-off-by: Qiming Yang <qiming.yang@intel.com>
Acked-by: Remy Horton <remy.horton@intel.com>

net/virtio: fix advertised Rx offload capabilities

When the virtio PMD is used on top of a vhost that does not support
offloads, Rx offload capabilities are still advertised by
virtio_dev_info_get(). But if an application tries to start the PMD with
Rx offloads enabled (rxmode.hw_ip_checksum = 1), the initialization of
the device will fail with -ENOTSUP and the following log:

  rx ip checksum not available on this host

This patch fixes the Rx offload capabilities returned by
virtio_dev_info_get() to be consistent with features advertised by the
host.

Fixes: 96cb6711939e ("net/virtio: support Rx checksum offload")
Fixes: 86d59b21468a ("net/virtio: support LRO")
Cc: stable@dpdk.org
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

examples/performance-thread: add packet type parsing

Last changes in Niantic and Fortville NIC drivers causes that
vector Rx path is chosen by default in l3fwd-thread application.
This path doesn't support propagation of hw packet type recognition
to the packet_type field in mbuf, and packets cannot be classified
properly.

The approach to solve this problem is similar to the commit:
71a7e2424e07 ("examples/l3fwd: fix using packet type blindly").

To use sw packet analyzer, new command line option "--parse-ptype" is
introduced.

Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com>

examples/ip_pipeline: fix parsing of pass-through pipeline

This patch fixes the configuration file parsing error when load balancing
function is enabled in pass-through pipeline.

error log:
pipeline> [APP] Initializing PIPELINE1 ...
[PIPELINE1] Pass-through
Parse error in section "PIPELINE1": entry "lb" has invalid value ("hash")

Fixes: cbe82f6cfb0a ("examples/ip_pipeline: add swap action in pass-through")
Cc: stable@dpdk.org
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>

examples/ip_pipeline: fix coremask limitation

Issue:
coremask used in IP Pipeline is limited to 64 cores.

Solution:
Modified coremask as an array of uint64_t to support RTE_MAX_LCORE

Fixes: 7f64b9c004aa ("examples/ip_pipeline: rework config file syntax")
Fixes: eb32fe7c5574 ("examples/ip_pipeline: rework initialization parameters")
Fixes: b4aee0fb9c6d ("examples/ip_pipeline: reconfigure thread binding dynamically")
Fixes: 4e14069328fc ("examples/ip_pipeline: measure CPU utilization")

Signed-off-by: Sankar Chokkalingam <sankarx.chokkalingam@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>

examples/ip_pipeline: check VLAN and MPLS parameters

This commit add to CLI command check for the following errors
1. SVLAN and CVLAN IDs greater than 12 bits
2. MPLS ID greater than 20 bits
3. max number of supported MPLS labels to avoid array overflow

It prevents running CLI commands with invalid parameters.

Signed-off-by: Anand B Jyoti <anand.b.jyoti@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>

examples/ip_pipeline: remove useless makefile line

A dollar sign is missing and it is not needed because of VPATH.

Reported-by: Ilya V. Matveychikov <matvejchikov@gmail.com>
Suggested-by: Ferruh Yigit <ferruh.yigit@intel.com>
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>

examples/l3fwd: rework long options parsing

Avoid the use of several strncpy() since getopt is able to
map a long option with an id, which can be matched in the
same switch/case than short options.

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

examples/l2fwd: rework long options parsing

Do the same than in l3fwd to avoid strcmp() for long options.

For l2fwd, there is no long option that take advantage of this new
mechanism as --mac-updating and --no-mac-updating are directly setting a
flag without needing an entry in the switch/case.

So this patch just prepares the framework in case a new long option is
added in the future.

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

doc: fix links to Linux in contribution guide

A referenced document in the Linux Kernel has been moved to a
sub-directory. And kernel community has moved to RST/Sphinx. The links are
replaced with HTML rendered links.

Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

doc: simplify l3fwd example guide

L3 Forwarding sample app user guides have some inconsistencies
between the example command line and the configuration table.
Also, they were showing too complicated configuration, using two
different NUMA nodes for two ports, which will probably lead
to performance drop due to use cross-socket channel.

This patch simplifies the configuration of these examples,
by using a single NUMA node and a single queue per port.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

doc: fix a typo in prog guide

Signed-off-by: Yong Wang <wang.yong19@zte.com.cn>
Acked-by: John McNamara <john.mcnamara@intel.com>

doc: fix a typo in proc_info guide

This patch fixes a typo in proc_info guide (tools).

Signed-off-by: Rami Rosen <rami.rosen@intel.com>

doc: fix a typo in testpmd guide

This patch fixes a trivial typo in testpmd application guide.

Signed-off-by: Rami Rosen <rami.rosen@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

app/testpmd: fix check for invalid ports

Some CLIs don't check the input port ID, it
may cause segmentation fault (core dumped).

Fixes: 425781ff5afe ("app/testpmd: add ixgbe VF management")

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Signed-off-by: Chen Jing D(Mark) <jing.d.chen@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>

eal: optimize aligned memcpy on x86

This patch optimizes rte_memcpy for well aligned cases, where both
dst and src addr are aligned to maximum MOV width. It introduces a
dedicated function called rte_memcpy_aligned to handle the aligned
cases with simplified instruction stream. The existing rte_memcpy
is renamed as rte_memcpy_generic. The selection between them 2 is
done at the entry of rte_memcpy.

The existing rte_memcpy is for generic cases, it handles unaligned
copies and make store aligned, it even makes load aligned for micro
architectures like Ivy Bridge. However alignment handling comes at
a price: It adds extra load/store instructions, which can cause
complications sometime.

DPDK Vhost memcpy with Mergeable Rx Buffer feature as an example:
The copy is aligned, and remote, and there is header write along
which is also remote. In this case the memcpy instruction stream
should be simplified, to reduce extra load/store, therefore reduce
the probability of load/store buffer full caused pipeline stall, to
let the actual memcpy instructions be issued and let H/W prefetcher
goes to work as early as possible.

This patch is tested on Ivy Bridge, Haswell and Skylake, it provides
up to 20% gain for Virtio Vhost PVP traffic, with packet size ranging
from 64 to 1500 bytes.

The test can also be conducted without NIC, by setting loopback
traffic between Virtio and Vhost. For example, modify the macro
TXONLY_DEF_PACKET_LEN to the requested packet size in testpmd.h,
rebuild and start testpmd in both host and guest, then "start" on
one side and "start tx_first 32" on the other.

Signed-off-by: Zhihong Wang <zhihong.wang@intel.com>
Reviewed-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>

examples/l3fwd-power: fix stop and close on signal

As it gets killed, in SIGINT signal handler, device is not stopped
and closed. In virtio's case, vector assignment in the KVM is not
deassigned.

This patch will invoke dev_stop() and dev_close() in signal handler.

Fixes: d7937e2e3d12 ("power: initial import")

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>

examples/l3fwd-power: add --parse-ptype option

To support those devices that do not provide packet type info when
receiving packets, add a new option, --parse-ptype, to analyze
packet type in the Rx callback.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>

net/virtio: unmap queue/irq when closing

When closing virtio devices, close eventfds, free the struct to
store queue/irq mapping.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio: unbind interrupt/eventfd when stopping

When virtio devices get stopped, tell the kernel to unbind the
mapping between interrupts and eventfds.

Note: it behaves differently from other NICs which close eventfds,
free struct. In virtio, we do those things when close device in
following patch.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio: setup Rx queue interrupts

This patch mainly allocates structure to store queue/irq mapping,
and configure queue/irq mapping down through PCI ops. It also creates
eventfds for each Rx queue and tell the kernel about the eventfd/intr
binding.

Note: So far, we hard-code 1:1 queue/irq mapping (each rx queue has
one exclusive interrupt), like this:
  vec 0 -> config irq
  vec 1 -> rxq0
  vec 2 -> rxq1
  ...

which means, the "vectors" option of QEMU should be configured with
a value >= N+1 (N is the number of the queue pairs).

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio: add Rx interrupt enable/disable functions

This patch implements interrupt enable/disable functions for each
Rx queue. And we rely on flags of avail queue as the hint for virtio
device to interrupt virtio driver or not.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio: add PCI operation for queue/irq binding

Add handler in virtio_pci_ops to set queue/irq bind.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio: add Rx descriptor check

Under interrupt mode, rx_descriptor_done is used as an indicator
for applications to check if some number of packets are ready to
be received.

This patch enables this by checking used ring's local consumed idx
with shared (with backend) idx.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio: invoke method directly for setting IRQ config

We need to define a prototype for such wrapper, which makes thing
too complicated. Remove wrapper and call set_config_irq directly.

Suggested-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio: fix rewriting LSC flag

The LSC flag is decided according to if VIRTIO_NET_F_STATUS feature
is negotiated. Copy the PCI info after the judgement will rewrite
the correct result.

Fixes: 198ab33677c9 ("net/virtio: move device initialization in a function")
CC: stable@dpdk.org
Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Tested-by: Lei Yao <lei.a.yao@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio-user: enable multiqueue with kernel vhost

With vhost kernel, to enable multiqueue, we need backend device
in kernel support multiqueue feature. Specifically, with tap
as the backend, as linux/Documentation/networking/tuntap.txt shows,
we check if tap supports IFF_MULTI_QUEUE feature.

And for vhost kernel, each queue pair has a vhost fd, and with a tap
fd binding this vhost fd. All tap fds are set with the same tap
interface name.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio-user: enable offloading

When used with vhost kernel backend, we can offload at both directions.
  - From vhost kernel to virtio_user, the offload is enabled so that
    DPDK app can trust the flow is checksum-correct; and if DPDK app
    sends it through another port, the checksum needs to be
    recalculated or offloaded. It also applies to TSO.
  - From virtio_user to vhost_kernel, the offload is enabled so that
    kernel can trust the flow is L4-checksum-correct, no need to verify
    it; if kernel will consume it, DPDK app should make sure the
    l3-checksum is correctly set.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio-user: support kernel vhost

This patch add support vhost kernel as the backend for virtio_user.
Three main hook functions are added:
  - vhost_kernel_setup() to open char device, each vq pair needs one
    vhostfd;
  - vhost_kernel_ioctl() to communicate control messages with vhost
    kernel module;
  - vhost_kernel_enable_queue_pair() to open tap device and set it
    as the backend of corresonding vhost fd (that is to say, vq pair).

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio-user: abstract backend operations

Add a struct virtio_user_backend_ops to abstract three kinds of backend
operations:
  - setup, create the unix socket connection;
  - send_request, sync messages with backend;
  - enable_qp, enable some queue pair.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net/virtio-user: move vhost-user specific code

To support vhost kernel as the backend of net_virtio_user in coming
patches, we move vhost_user specific structs and macros into
vhost_user.c, and only keep common definitions in vhost.h.

Besides, remove VHOST_USER_MQ feature check.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>